SUMMARY Project 3 will develop new computational tools for understanding how gut microbial communities change their membership, gene content, gene expression, and metabolic activities in obesity and during diet interventions, in both humans and 'humanized' gnotobiotic mouse models. Its overarching goals are to: (i) understand the levels at which multi-omics data should be collected and analyzed in order to maximize our understanding of complex multifactorial pathophysiological conditions such as obesity and its associated metabolic abnormalities; (ii) develop improved genome assembly techniques and predictions about culture conditions and syntrophic interactions to improve the utility of personalized bacterial culture collections and data derived from them; and (iii) understand how best to characterize the diversity of gut microbial communities, and the functional profiles of these communities, observed in the human population and use them for patient stratification. Project 3 has three Aims: (1) develop new tools for relating multi-omics data across analysis levels and relating information from mouse models, specifically gnotobiotic humanized mice characterized in Project 1 and Core A, to information about the discordant twins (phenotyped in Project 2 and Core A) from which those animal models were derived and to the human population at large; (2) to develop improved methods for assembly of complete bacterial genomes as a reference for shotgun metagenomic and meta- transcriptomic data, including meta-transcriptome data collected from gnotobiotic mice colonized with bacterial culture collections generated from the fecal microbiota of co-twins in discordant twin pairs where the complete bacterial genomes are known, thus eliminating a major computational bottleneck and providing more relevant types of assemblies for downstream annotation and interpretation tasks; and (3) to provide a broader understanding of the major patterns of variation in human gut microbial communities and their genes, transcripts and metabolites in individuals with and without obesity and obesity associated abnormalities, and to test whether these major patterns can be used for stratification of human subjects in terms of their response to specific dietary or other interventions. Project 3 and Core B will work closely together to make new analysis tools and datasets available to the scientific community.